Using the quality of health economic studies instrument to assess pharmacoeconomic studies evaluating recent food and drug administration drug approvals

To determine study quality, estimate quality scoring reliability, and assess theorized quality predictors of pharmacoeconomic publications evaluating recent new molecular entity and biologic license approvals (NMEs) by the Food and Drug Administration (FDA). Original pharmacoeconomic studies (cost-effectiveness, cost-utility, cost-benefit or cost-minimization) considering any of 50 NMEs approved in 2008-09 and published on or before December 31, 2011 were eligible. MEDLINE and the UK National Health Service Economic Evaluation Database were searched. Quality was scored with the Quality of Health Economic Studies (QHES) instrument for each publication by one primary and two secondary reviewers. Interrater reliability was assessed using Pearson correlations of QHES scores. Regression was performed of QHES score on study characteristics including number of authors, journal impact factor one-year pre-submission, journal type (disease-specific/general clinical/health economic), NME FDA review classifications (priority/standard and orphan/non), publication timing (pre-/post-NME approval), author(s) having academic affiliation (yes/no), advanced modeling PE techniques (yes/no), United States study (yes/no), data (primary/secondary), incorporation of quality-adjusted life years (yes/no), and conclusion (favorable/unfavorable for NME). The literature search yielded 203 search results with 37 publications meeting inclusion criteria, encompassing 38% of the 2008-09 NMEs. Averaging all reviewers, the QHES score range was 15-92, with a median 70, and mean 68.4±18.4. The total QHES score was significantly correlated between reviewers (R= 0.677). A square transformation was applied to QHES score to correct for a negatively skewed distribution. Regression analyses were non-significant for all study characteristics, although use of advanced modeling PE techniques approached significance (p=0.083). QHES scores indicated that the quality of pharmacoeconomic literature for newly-approved NMEs varies, although the 51.3%, 19, highest scoring studies (including and above the median) were near or exceeded the 75 point threshold considered “good”. Interrater reliability for QHES assessment was fair. Sample size was insufficient to identify significant predictors among the variables analyzed.